Institute of Mechanics,
Chinese Academy of Sciences
2024 Vol.14(1)
Display Mode： |
Theoretical and Applied Mechanics Letters 14 (2024) 100491.
doi: 10.1016/j.taml.2024.100491
Abstract:
Reinforcement learning (RL) algorithms are expected to become the next generation of wind farm control methods. However, as wind farms continue to grow in size, the computational complexity of collective wind farm control will exponentially increase with the growth of action and state spaces, limiting its potential in practical applications. In this Letter, we employ a RL-based wind farm control approach with multi-agent deep deterministic policy gradient to optimize the yaw manoeuvre of grouped wind turbines in wind farms. To reduce the computational complexity, the turbines in the wind farm are grouped according to the strength of the wake interaction. Meanwhile, to improve the control efficiency, each subgroup is treated as a whole and controlled by a single agent. Optimized results show that the proposed method can not only increase the power production of the wind farm but also significantly improve the control efficiency.
Reinforcement learning (RL) algorithms are expected to become the next generation of wind farm control methods. However, as wind farms continue to grow in size, the computational complexity of collective wind farm control will exponentially increase with the growth of action and state spaces, limiting its potential in practical applications. In this Letter, we employ a RL-based wind farm control approach with multi-agent deep deterministic policy gradient to optimize the yaw manoeuvre of grouped wind turbines in wind farms. To reduce the computational complexity, the turbines in the wind farm are grouped according to the strength of the wake interaction. Meanwhile, to improve the control efficiency, each subgroup is treated as a whole and controlled by a single agent. Optimized results show that the proposed method can not only increase the power production of the wind farm but also significantly improve the control efficiency.
Theoretical and Applied Mechanics Letters 14 (2024) 100488.
doi: 10.1016/j.taml.2023.100488
Abstract:
With the increased availability of experimental measurements aiming at probing wind resources and wind turbine operations, machine learning (ML) models are poised to advance our understanding of the physics underpinning the interaction between the atmospheric boundary layer and wind turbine arrays, the generated wakes and their interactions, and wind energy harvesting. However, the majority of the existing ML models for predicting wind turbine wakes merely recreate CFD-simulated data with analogous accuracy but reduced computational costs, thus providing surrogate models rather than enhanced data-enabled physics insights. Although ML-based surrogate models are useful to overcome current limitations associated with the high computational costs of CFD models, using ML to unveil processes from experimental data or enhance modeling capabilities is deemed a potential research direction to pursue. In this letter, we discuss recent achievements in the realm of ML modeling of wind turbine wakes and operations, along with new promising research strategies.
With the increased availability of experimental measurements aiming at probing wind resources and wind turbine operations, machine learning (ML) models are poised to advance our understanding of the physics underpinning the interaction between the atmospheric boundary layer and wind turbine arrays, the generated wakes and their interactions, and wind energy harvesting. However, the majority of the existing ML models for predicting wind turbine wakes merely recreate CFD-simulated data with analogous accuracy but reduced computational costs, thus providing surrogate models rather than enhanced data-enabled physics insights. Although ML-based surrogate models are useful to overcome current limitations associated with the high computational costs of CFD models, using ML to unveil processes from experimental data or enhance modeling capabilities is deemed a potential research direction to pursue. In this letter, we discuss recent achievements in the realm of ML modeling of wind turbine wakes and operations, along with new promising research strategies.
Theoretical and Applied Mechanics Letters 14 (2024) 100494.
doi: 10.1016/j.taml.2024.100494
Abstract:
Physics-informed neural networks (PINN) are a useful machine learning method for solving differential equations, but encounter challenges in effectively learning thin boundary layers within singular perturbation problems. To resolve this issue, Multi-Scale-Matching Neural Networks (MSM-NN) are proposed to solve the singular perturbation problems. Inspired by matched asymptotic expansions, the solution is decomposed into inner solutions for small scales and outer solutions for large scales, corresponding to boundary layers and outer regions, respectively. Moreover, to conform neural networks, we introduce exponential stretched variables in the boundary layers to avoid semi-infinite region problems. Numerical results for the thin plate problem validate the proposed method.
Physics-informed neural networks (PINN) are a useful machine learning method for solving differential equations, but encounter challenges in effectively learning thin boundary layers within singular perturbation problems. To resolve this issue, Multi-Scale-Matching Neural Networks (MSM-NN) are proposed to solve the singular perturbation problems. Inspired by matched asymptotic expansions, the solution is decomposed into inner solutions for small scales and outer solutions for large scales, corresponding to boundary layers and outer regions, respectively. Moreover, to conform neural networks, we introduce exponential stretched variables in the boundary layers to avoid semi-infinite region problems. Numerical results for the thin plate problem validate the proposed method.
Theoretical and Applied Mechanics Letters 14 (2024) 100486.
doi: 10.1016/j.taml.2023.100486
Abstract:
Besides showing excellent abilities such as energy absorption, phase-transforming metamaterials provide a rich design space for achieving nonlinear constitutive relations by switching between different patterns under deformation. The related inverse design problem, nevertheless, is quite challenging due to the lack of appropriate mathematical formulation and the convergence issue of post-buckling analysis of intermediate designs. In the present work, periodic unit cells are explicitly described by the moving morphable voids method and effectively analyzed by removing the DOFs of void regions. Furthermore, exploring the Pareto frontiers between error and cost, an inverse design formulation is proposed for unit cells achieving a prescribed constitutive curve and validated by numerical examples and experimental results. The present design approach can be extended to the inverse design of other types of mechanical metamaterials with prescribed nonlinear effective properties.
Besides showing excellent abilities such as energy absorption, phase-transforming metamaterials provide a rich design space for achieving nonlinear constitutive relations by switching between different patterns under deformation. The related inverse design problem, nevertheless, is quite challenging due to the lack of appropriate mathematical formulation and the convergence issue of post-buckling analysis of intermediate designs. In the present work, periodic unit cells are explicitly described by the moving morphable voids method and effectively analyzed by removing the DOFs of void regions. Furthermore, exploring the Pareto frontiers between error and cost, an inverse design formulation is proposed for unit cells achieving a prescribed constitutive curve and validated by numerical examples and experimental results. The present design approach can be extended to the inverse design of other types of mechanical metamaterials with prescribed nonlinear effective properties.
Theoretical and Applied Mechanics Letters 14 (2024) 100493.
doi: 10.1016/j.taml.2024.100493
Abstract:
In recent years, materials with asymmetric mechanical response properties (mechanical Janus materials) have been found possess numerous potential applications， i.e. shock absorption and vibration isolation. In this study, we propose a novel mechanical Janus lattice whose asymmetric mechanical response can be switched in orientation by a plug. Through finite element analysis (FEA) and experimental verification, this lattice exhibits asymmetric displacement responses to symmetric forces. Furthermore, with such a plug structure inside, individual lattices can switch the orientation of asymmetry and thus achieve reprogrammable design of a mechanical structure with chained lattices. The reprogrammable asymmetry of this material will offer multiple functions in design of mechanical metamaterials
In recent years, materials with asymmetric mechanical response properties (mechanical Janus materials) have been found possess numerous potential applications， i.e. shock absorption and vibration isolation. In this study, we propose a novel mechanical Janus lattice whose asymmetric mechanical response can be switched in orientation by a plug. Through finite element analysis (FEA) and experimental verification, this lattice exhibits asymmetric displacement responses to symmetric forces. Furthermore, with such a plug structure inside, individual lattices can switch the orientation of asymmetry and thus achieve reprogrammable design of a mechanical structure with chained lattices. The reprogrammable asymmetry of this material will offer multiple functions in design of mechanical metamaterials
Theoretical and Applied Mechanics Letters 14 (2024) 100489.
doi: 10.1016/j.taml.2023.100489
Abstract:
Conventional wing aerodynamic optimization processes can be time-consuming and imprecise due to the complexity of versatile flight missions. Plenty of existing literature has considered twodimensional infinite airfoil optimization, while three-dimensional finite wing optimizations are subject to limited study because of high computational costs. Here we create an adaptive optimization methodology built upon a digitized wing shape deformation and deep learning algorithms, which enable the rapid formulation of finite wing designs for specific aerodynamic performance demands under different cruise conditions. This methodology unfolds in three stages: radial basis function (RBF) interpolated wing generation, collection of inputs from computational fluid dynamics (CFD) simulations, and deep neural network that constructs the surrogate model for the optimal wing configuration. To this end, the proposed methodology can significantly reduce the computational cost of numerical simulations. It also has potential to optimize various aerial vehicles undergoing different mission environments, loading conditions, and safety requirements.
Conventional wing aerodynamic optimization processes can be time-consuming and imprecise due to the complexity of versatile flight missions. Plenty of existing literature has considered twodimensional infinite airfoil optimization, while three-dimensional finite wing optimizations are subject to limited study because of high computational costs. Here we create an adaptive optimization methodology built upon a digitized wing shape deformation and deep learning algorithms, which enable the rapid formulation of finite wing designs for specific aerodynamic performance demands under different cruise conditions. This methodology unfolds in three stages: radial basis function (RBF) interpolated wing generation, collection of inputs from computational fluid dynamics (CFD) simulations, and deep neural network that constructs the surrogate model for the optimal wing configuration. To this end, the proposed methodology can significantly reduce the computational cost of numerical simulations. It also has potential to optimize various aerial vehicles undergoing different mission environments, loading conditions, and safety requirements.
Theoretical and Applied Mechanics Letters 14 (2024) 100483.
doi: 10.1016/j.taml.2023.100483
Abstract:
Bistate plates have found extensive applications in the domains of smart structures and energy harvesting devices. Most bistable curved plates are characterized by a constant thickness profile. Regrettably, due to the inherent complexity of this problem, relatively little attention has been devoted to this area. In this study, we demonstrate how deep learning can facilitate the discovery of novel plate profiles that cater to multiple objectives, including maximizing stiffness, forward snapping force, and backward snapping force. Our proposed approach is distinguished by its efficiency in terms of low computational energy consumption and high effectiveness. It holds promise for future applications in the design and optimization of multistable structures with diverse objectives, addressing the requirements of various fields.
Bistate plates have found extensive applications in the domains of smart structures and energy harvesting devices. Most bistable curved plates are characterized by a constant thickness profile. Regrettably, due to the inherent complexity of this problem, relatively little attention has been devoted to this area. In this study, we demonstrate how deep learning can facilitate the discovery of novel plate profiles that cater to multiple objectives, including maximizing stiffness, forward snapping force, and backward snapping force. Our proposed approach is distinguished by its efficiency in terms of low computational energy consumption and high effectiveness. It holds promise for future applications in the design and optimization of multistable structures with diverse objectives, addressing the requirements of various fields.
Theoretical and Applied Mechanics Letters 14 (2024) 100484.
doi: 10.1016/j.taml.2023.100484
Abstract:
Discrete dislocation dynamics (DDD) simulations reveal the evolution of dislocation structures and the interaction of dislocations. This study investigated the compression behavior of single-crystal copper micropillars using few-shot machine learning with data provided by DDD simulations. Two types of features are considered: external features comprising specimen size and loading orientation and internal features involving dislocation source length, Schmid factor, the orientation of the most easily activated dislocations and their distance from the free boundary. The yielding stress and stress-strain curves of single-crystal copper micropillar are predicted well by incorporating both external and internal features of the sample as separate or combined inputs. It is found that the Machine learning accuracy predictions for single-crystal micropillar compression can be improved by incorporating easily activated dislocation features with external features. However, the effect of easily activated dislocation on yielding is less important compared to the effects of specimen size and Schmid factor which includes information of orientation but becomes more evident in small-sized micropillars. Overall, incorporating internal features, especially the information of most easily activated dislocations, improves predictive capabilities across diverse sample sizes and orientations.
Discrete dislocation dynamics (DDD) simulations reveal the evolution of dislocation structures and the interaction of dislocations. This study investigated the compression behavior of single-crystal copper micropillars using few-shot machine learning with data provided by DDD simulations. Two types of features are considered: external features comprising specimen size and loading orientation and internal features involving dislocation source length, Schmid factor, the orientation of the most easily activated dislocations and their distance from the free boundary. The yielding stress and stress-strain curves of single-crystal copper micropillar are predicted well by incorporating both external and internal features of the sample as separate or combined inputs. It is found that the Machine learning accuracy predictions for single-crystal micropillar compression can be improved by incorporating easily activated dislocation features with external features. However, the effect of easily activated dislocation on yielding is less important compared to the effects of specimen size and Schmid factor which includes information of orientation but becomes more evident in small-sized micropillars. Overall, incorporating internal features, especially the information of most easily activated dislocations, improves predictive capabilities across diverse sample sizes and orientations.
Theoretical and Applied Mechanics Letters 14 (2024) 100490.
doi: 10.1016/j.taml.2024.100490
Abstract:
Dielectric elastomers (DEs) require balanced electric actuation performance and mechanical integrity under applied voltages. Incorporating high dielectric particles as fillers provides extensive design space to optimize concentration, morphology, and distribution for improved actuation performance and material modulus. This study presents an integrated framework combining finite element modeling (FEM) and deep learning to optimize the microstructure of DE composites. FEM first calculates actuation performance and the effective modulus across varied filler combinations, with these data used to train a convolutional neural network (CNN). Integrating the CNN into a multi-objective genetic algorithm (NSGA-II) generates designs with enhanced actuation performance and material modulus compared to the conventional FEM-NSGA-II approach within the same time. This framework harnesses artificial intelligence to navigate vast design possibilities, enabling optimized microstructures for high-performance DE composites.
Dielectric elastomers (DEs) require balanced electric actuation performance and mechanical integrity under applied voltages. Incorporating high dielectric particles as fillers provides extensive design space to optimize concentration, morphology, and distribution for improved actuation performance and material modulus. This study presents an integrated framework combining finite element modeling (FEM) and deep learning to optimize the microstructure of DE composites. FEM first calculates actuation performance and the effective modulus across varied filler combinations, with these data used to train a convolutional neural network (CNN). Integrating the CNN into a multi-objective genetic algorithm (NSGA-II) generates designs with enhanced actuation performance and material modulus compared to the conventional FEM-NSGA-II approach within the same time. This framework harnesses artificial intelligence to navigate vast design possibilities, enabling optimized microstructures for high-performance DE composites.
Theoretical and Applied Mechanics Letters 14 (2024) 100492.
doi: 10.1016/j.taml.2024.100492
Abstract:
The distribution of material phases is crucial to determine the composite’s mechanical property. While the full structure-mechanics relationship of highly ordered material distributions can be studied with finite number of cases, this relationship is difficult to be revealed for complex irregular distributions, preventing design of such material structures to meet certain mechanical requirements. The noticeable developments of artificial intelligence (AI) algorithms in material design enables to detect the hidden structuremechanics correlations which is essential for designing composite of complex structures. It is intriguing how these tools can assist composite design. Here, we focus on the rapid generation of bicontinuous composite structures together with the stress distribution in loading. We find that generative AI, enabled through fine-tuned Low Rank Adaptation models, can be trained with a few inputs to generate both synthetic composite structures and the corresponding von Mises stress distribution. The results show that this technique is convenient in generating massive composites designs with useful mechanical information that dictate stiffness, fracture and robustness of the material with one model, and such has to be done by several different experimental or simulation tests. This research offers valuable insights for the improvement of composite design with the goal of expanding the design space and automatic screening of composite designs for improved mechanical functions.
The distribution of material phases is crucial to determine the composite’s mechanical property. While the full structure-mechanics relationship of highly ordered material distributions can be studied with finite number of cases, this relationship is difficult to be revealed for complex irregular distributions, preventing design of such material structures to meet certain mechanical requirements. The noticeable developments of artificial intelligence (AI) algorithms in material design enables to detect the hidden structuremechanics correlations which is essential for designing composite of complex structures. It is intriguing how these tools can assist composite design. Here, we focus on the rapid generation of bicontinuous composite structures together with the stress distribution in loading. We find that generative AI, enabled through fine-tuned Low Rank Adaptation models, can be trained with a few inputs to generate both synthetic composite structures and the corresponding von Mises stress distribution. The results show that this technique is convenient in generating massive composites designs with useful mechanical information that dictate stiffness, fracture and robustness of the material with one model, and such has to be done by several different experimental or simulation tests. This research offers valuable insights for the improvement of composite design with the goal of expanding the design space and automatic screening of composite designs for improved mechanical functions.
Submit a Paper
Subscription
News
MORE+
Call for Papers
MORE+
- Crossing-Mechanics Driven by Big Data
- Machine learning in the fluid mechanics research of wind energy
- Mechanics of Origami/Kirigami structures and metamaterials
- New insights and perspectives on impact biomechanics for human tissues: from injury prevention, protection to protective equipment
- Environmental Mechanics for Extreme Natural Events
